An Article by Samaksh Khanna.
How often have we observed that every billion dollar idea starts with an algorithm on the window. An algorithm can decide the outcome or income of money into the financial sectors. An investor today heavily relies on the functioning of the economy and this can be determined by various forms of statistical analysis. Let’s take the market downfall of 2008 as an example. Nobody was ready to believe that the entire housing economy in the United States could come to a standstill, but it did happen. It was predicted by a few people on Wall Street through statistics and mathematical algorithms but nobody paid any attention to them. The power of intuition combined with the experience of the scenarios of the market is what banks and stock market trader have been operating on, during the past years. These traditional systems are known as fundamental ways of investing, however, a new verge of trading platforms seems to have emerged in the financial sector which is called ‘quantamental investing.’
Quantamental investing is defined as the future of the asset management industry. It is an innovate method of investing which combines two terms, fundamental and quantitative. The quantitative (or quant) method of investing uses higher end business models and algorithms to suggest the flow of the market. While the fundamental method is an old school method which is based on intuition and keeping a track of industry trends.
Quantamental investing is a more disciplined and a direct process. Its main objective is to evaluate the datasets and assess the risk in the prevailing market conditions. This helps in improving judgment and helps in monitoring the entire investment process. [1] One of the most important techniques is to build statistical models in order to manage risk and allocate funds in a better way thus ensuring that not much is actually left to chance. Many asset management firms are moving towards strategies that rely more on algorithm and models rather than human intuition to pick stocks.
Companies such as Bridgewater and Goldman Sachs are testing the waters of quantamental investment strategies, WorldQuant is using artificial intelligence (“AI”) for small-scale trading, and Sentient Technologies conducts all its trading using AI.[2]
JP Morgan’s Asset Management’s chief data scientist says that “we do not come up with strategies out of thin air, there’s stuff that happens in the human brain that is so hard to replicate”. In light of this, JP Morgan’s 1.7 trillion dollars investment arm, in January 2018, set up a new data lab in its intelligent digital solutions division to try to improve its portfolio managers.[3]
JP Morgan is, in fact, doing something very unique. They are analysing how markets work and trying to establish the trends by developing ‘transcripts’. These transcripts analyse words which are particularly sensitive for markets or the words that may cause some kind of trouble. Let’s take for an example the word ‘great’. It has been observed that the usage of the word ‘great’ deems to be effective and profitable for the stock market. The word ‘debt’ however may be bad and may bring about some losses. Now a human manager or a trader at a time may be able to read a dozen transcripts at a time, however, a machine can scour thousands. The data analytics system senses the tone of the market and immediately sends an alert so that analysts can immediately examine their investment pattern. Some investment groups are starting to use technology to spot well-known behavioural biases. For example, Essentia Analytics crunches individual trading data and looks for common weak spots, such as fund managers’ tendency to over-trade when on a losing streak, or hang on to poor investments for too long to avoid crystallising losses. When that happens, fund managers get sent an automated but personalised email signed ‘your future self’ reminding them to be aware of these pitfalls.
Quantamental investing may sound a very mechanical method but the concept is not entirely new. Human analysis and data have been combined in various instances before and can be dated back to World War II. Dr Soresen suggests that whoever has a doubt or is unsure of quantamental investing may only look back to Second World War when Air Chief Marshal Hugh Dowding built a system of integrated information using radar, human observations and a dedicated phone system to overcome an enemy with roughly equal firepower.[5] This versatile combination allows stock managers to generate stronger returns over a broader stock universe.
It is pertinent to note that the practicality of using the quantamental method is easier said than done. While analysts are technophobes and love the numbers game, they tend to grow weary of computer-driven data and often insist that human judgment cannot be replicated. Alan Bochman, a partner in New York-based Genpact’s capital-markets consulting practice, says “the cultural issue is really, really hard to get through, and the way that it comes across to us is people from a fundamental background are very sceptical of automated solutions”.[6]
An important question that comes into mind is when algorithms are managing the money, who is actually managing all these algorithms? Markets are highly unpredictable in nature and thus it is necessary for some human intervention for picking stocks. The main skillset that humans can bring into the market which quantitative analysis cannot is the ability to perform when the information is based on small data sets. There are various instances where humans can do a better job especially when there is a factor of low risk, meaning that the stocks may have low expected returns. Now obviously when doing a stock analysis of the top global equities, the use of the algorithm would be more efficient.
What now has to be understood is how much risk is actually involved while asserting the quant analysis. It has already been stated that there are few instances where human analysis is more beneficial rather than algorithm data. Markets are uncertain, especially when interfered with political intervention. Securities can be volatile at times and it is not necessary that particular security performs the way it had been anticipated. One of the most important factors to note is that quant analysis does not take a day to day life aspects into consideration. It is only virtual and looks into the course of the market. Sometimes there can be natural calamities and this may result in the collapse of the market, a consideration which cannot be completely anticipated through quant analysis. Even during the times of an election, markets tend to perform in a very unexpected manner, and looking into the daily analysis of quants, knowing that they might change within a matter of hours is not very feasible.
To conclude it is important to say that quantamental investment is a big leap and will play a major role in the future of financial investing. But even a good quant trading strategy requires a good fundamental behind it. As more and more firms are developing quantamental techniques, it is likely that in the next 10 to 20 years the employment rate of analysts would fall sharply as productivity would increase. Computers can perform calculations of stocks and correlations better than human but not all strategies work in every considerable time.
[2] Adam Satariano, “Silicon Valley hedge fund takes on Wall Street with AI trader,” Bloomberg, February 6, 2017, bloomberg.com.
[3] Wiggelsworth, Robin. The rise of quantamental investing: where man and machine meet. The Financial Times, 21 November 2018.
[4] Wiggelsworth, Robin. The rise of quantamental investing: where man and machine meet. The Financial Times, 21 November 2018.
Comments